No. 011 September 28, 2026 9 min read

Four Parts in Ten Thousand

Qatar's North Field is 0.04 percent helium and supplies roughly a third of the world's. That only works because someone else is already paying to chill the gas — which is why a March strike on an LNG terminal took helium off the market that no price can bring back.

Qatar’s North Field, the largest single accumulation of natural gas on Earth, is 0.04 percent helium. QatarEnergy says so on its own operations page, in a sentence that reads like a footnote. Four parts in ten thousand. And that field, through the plants at Ras Laffan, supplies somewhere between a quarter and a third of the world’s helium.

I have been turning that number over for two days, because on its face it should not work. Helium at 0.04 percent is not an ore. The American fields that fed the federal helium programme for most of a century ran 0.3 to 2.7 percent — that is the USGS’s own range — and those are the rich ones — the reason Kansas and the Texas panhandle ended up holding a strategic reserve of a noble gas. Qatar’s gas is an order of magnitude leaner than the leanest of them, and is nonetheless the second-largest source on the planet.

0.04%

helium in Qatar's North Field

QatarEnergy's own figure

0.3–2.7%

helium in the US fields the federal reserve drew on

USGS

190 Mm³

world production, 2025

USGS estimate

The reason it works is that nobody is paying to get the helium out.

The separation is already bought and paid for

To ship gas as LNG you have to chill it to about −162 °C, and everything that will not condense at that temperature comes out first. What is left is a thin stream of nitrogen and whatever else refused to liquefy, and in it the helium — having started at four parts in ten thousand — is concentrated enough to be worth catching. The cryogenic plant that makes it recoverable was not built for helium. It was built because Japan and Korea and Europe want gas in a tanker.

This is the whole economic structure of the industry, and it has a consequence that took me a while to feel properly. Helium supply is not a response to helium demand. It is a residue of decisions made about something else. Anish Kapadia of AKAP Energy put it to AGBI in one line in March: “There is no helium being produced if LNG production is shut.” Not reduced. Not rationed. None.

Which is what happened. QatarEnergy declared force majeure on 3 March 2026 and shut liquefaction down; on 18 March the already-idle Ras Laffan complex was struck, and the damage — a reported 17 percent cut to Qatar’s LNG capacity, with repairs estimated at three to five years — turned a stoppage into a structural loss. Helium was not a target and not a consideration. It came off the market because a gas terminal stopped making a different product.

Two numbers that disagree, and both are right

The coverage I read gave two figures for how much of world helium supply went away. AGBI, in March, said 11 percent — about a third of Qatar’s expected 2026 volumes. The supply-chain firm Exiger said the shutdown took “approximately one-third of global helium supply” offline, working from roughly 5.2 million cubic metres a month removed.

Those look irreconcilable until you do the arithmetic. Qatar produced about 63 million cubic metres in 2025 out of a world total near 190 million — call it a third — and Exiger’s 5.2 million a month annualises to 62 million, which is all of Qatar. So Exiger is quoting a rate: while everything is down, a third of the world’s helium is not being made. AGBI is quoting a volume across calendar 2026, with a restart assumed partway through, which lands near a tenth. Both are defensible. They measure different things, and I watched both get repeated as the same kind of claim, usually in the same paragraph as a sentence about MRI machines.

A price that cannot summon anything

The USGS put the 2025 base price for Grade-A helium at about $330 per thousand cubic feet, and then added the quiet qualifier that producers post surcharges on top. That is a government statistical agency stating, in the price field, that it does not know the price. There is no exchange, no settlement, no published volume-weighted average. Trade reporting from distributors in mid-2026 has long-term contracts near $500–550 and spot as high as $1,000–1,200, which if true is a threefold move in a year.

Helium, US$ per thousand cubic feet

USGS base price, Grade-A, 2025

$330

Long-term contract, mid-2026 (trade reported)

~$500–550

Spot, mid-2026 (trade reported)

$1,000–1,200

Actual transacted average, any year

unreported

And here is the part I find genuinely interesting rather than merely alarming: that price increase cannot do the thing prices are supposed to do. A tripling of the copper price brings mothballed mines back. A tripling of the helium price brings back almost nothing, because the helium is not in anybody’s mine. It is in the tail gas of a liquefaction train whose owner is deciding whether to run based on the LNG price, and a helium windfall is a rounding error against that calculation.

There is a standalone helium industry — explorers like Helix Exploration and Blue Star Helium, drilling for the gas itself. Phil Kornbluth, who has watched this market for decades, told gasworld he does not “see primary helium accounting for more than 10% of the global supply.” The concentration arithmetic explains why. At 0.3 percent you need a reason to drill that is helium alone, and the entire US Grade-A helium business was worth about $970 million in 2025 against an LNG trade measured in hundreds of billions. The byproduct will always be able to outproduce the product.

The buffer that was sold, but not destroyed

The obvious thing to say next is that America sold its helium stockpile at exactly the wrong moment, and I nearly wrote that sentence before checking it.

The Federal Helium System went to Messer in June 2024 for $460 million, completing a wind-down Congress started in 1996 and scheduled in the Helium Stewardship Act of 2013. The reserve was not a token: in 2023 alone, 22.2 million cubic metres of crude helium were withdrawn from BLM storage, against total US sales of 81.2 million cubic metres. Call it a quarter of American supply coming out of a hole in the ground near Amarillo that had been filling since the Eisenhower administration, originally on the theory that the country would need it for airships.

But the gas did not evaporate when the deed changed hands, and neither did the cavern. Messer runs the same Bush Dome reservoir, the same pipeline, the same purification train. What ended was a public claim on the buffer — a statutory allocation, a price set by law rather than by a seller, and a federal obligation to hold something back. The physical stock survived privatisation; the policy instrument did not. When people say the US sold its helium and now there is a shortage, the causal story is wrong even though both halves are true, and the correct version is less cinematic and more interesting: the country stopped owning an option.

Five months

The detail that keeps catching me is the timing. In October 2025 the trade press was running features on helium oversupply — too much gas, softening prices, buyers renegotiating contracts downward. By late March 2026 the same publication was asking whether this was Helium Shortage 5.0. Nothing changed about the geology, and nothing much changed about demand. What changed was whether a particular set of cryogenic trains on the Persian Gulf were running.

The demand side offers less relief than the coverage suggests. Everyone reaches for MRI, and MRI is real — the USGS puts it at 15 percent of US consumption, while Exiger says healthcare is about 30 percent globally, a gap nobody seems interested in reconciling. Philips’s sealed BlueSeal magnets hold about 7 litres of liquid helium against roughly 1,500 for a conventional magnet, and the company claims some 2,000 systems installed and more than 6 million litres of helium saved since 2019. That is a genuinely good piece of engineering. It is also, converted to gas, about 4.5 million cubic metres — roughly 2 percent of a single year’s world production, accumulated over six years. The most celebrated efficiency win in the biggest sympathetic end-use does not move the supply problem.

Meanwhile the largest single line in US consumption, at 17 percent, is “lifting gas.” More than MRI. That category is weather balloons and airships as well as the ones at birthday parties, and the USGS does not break it out, which is convenient for everyone arguing about balloons in either direction.

What I am left with is a mental model I did not have on Friday. I had filed helium under scarce elements — a geology problem, a running-out problem, the sort of thing where the answer is conservation and recycling and eventually a price that fixes it. It is not that. The United States has 8.5 billion cubic metres in identified reserves, forty-odd years of world consumption, and Qatar’s are described simply as “large.” The gas is there. What is missing is any mechanism by which wanting helium causes helium to be produced. The price is a readout, not a control — an instrument that tells you how badly the residue of someone else’s business has been interrupted.

That is an unusual thing for a commodity to be, and I suspect it is more common than I thought. Sulphur, most of it a byproduct of desulphurising fuel. Rhenium out of copper. Tellurium, gallium, most of the world’s cobalt. There is a whole class of materials whose supply curve is not a curve at all, just a number handed down from a market nobody in the room is participating in. Helium is only the clearest case, because you cannot substitute it and you cannot mine it on purpose and the field that supplies a third of it is 99.96 percent something else.